Transaction behavior judgment method and system

Abstract

The invention relates to a transaction behavior judgment method and system. The transaction behavior judgment method comprises the following steps: generating sample data of transaction behaviors of known types; obtaining a transaction characteristic parameter set comprising transaction characteristic parameters for judging the transaction behavior type by using the sample data; and judging the unknown type of transaction behavior according to the associated historical transaction data of the unknown type of transaction behavior and the transaction characteristic parameters in the current transaction data. The transaction behavior judgment system comprises a sample data generation module configured to generate sample data of a known type of transaction behavior; a transaction characteristic parameter generation module configured to obtain a transaction characteristic parameter set comprising transaction characteristic parameters for judging transaction behavior types by using the sample data; and a transaction behavior type judgment module configured to judge the unknown type of transaction behavior according to the associated historical transaction data of the unknown type of transaction behavior and the transaction characteristic parameters in the current transaction data.

Description

Technical field [0001] The invention relates to a method and system for judging transaction behaviors. Background technique [0002] In the promotion of some products, discounts and other methods are usually used to market the products. At the same time, some people use preferential activities in product marketing to exchange material benefits at relatively low or zero costs. This part of the population is called the "scalper party." With the gradual development of product promotion business, different types of preferential activities are gradually increasing, and the behavior of using preferential activities to earn a price difference to obtain a large amount of benefits will seriously affect the promotion cost and marketing effect of the product. Due to the need for refined business operations, there is an urgent need to reasonably control such groups. [0003] The model based on rule judgment has been widely used in trading behavior detection. Although this method can reasonab...

AI Technical Summary

Problems solved by technology

[0003] The model based on rule judgment has been widely used in the detection of trading behavior. Although this method can reasonably detect and explain the trading behavior in the current time trading segment, it is based on empirical rule criteria and only for the current time. Fragment transaction behavior detection methods have been difficult to meet the needs of increasingly complex business scenarios

Lack of consideration of the impact of transaction time series, it is difficult to evaluate the user's historical transaction behavior, and it cannot cope with the ever-changing anti-detection methods

[0004] On the other hand, the abnormal transaction detection method based on the supervision framework has weak business explainability, and there is a great shortage in the acquisition of transaction behavior evidence

Moreover, the traditional rule-based scalper detection method mostly relies on manual identification. The rules and standards are determined through the common sense cognition of trading behavior, which is weak in adjustability and lacks the acquisition of trading rules and behavioral characteristics, which is not conducive to the expansion of scenarios and functions.

Images